This is a proposal to address statistical design and analysis issues related to the study of diabetes, especially in Hispanic populations. Ms. Arroyo is particularly interested in "community-based research" which provides a powerful approach to both understanding and intervening in the health concerns of underserved populations since it involves the community directly in study planning, design, implementation and interpretation. Although the specifics of her dissertation research will be developed over the next year or so, the kinds of issues that arise in this area include cluster randomization, longitudinal data analysis, missing data problems and functional data analysis.